i18n: make GETTEXT_POISON a runtime option

Change the GETTEXT_POISON compile-time + runtime GIT_GETTEXT_POISON
test parameter to only be a GIT_TEST_GETTEXT_POISON=<non-empty?>
runtime parameter, to be consistent with other parameters documented
in "Running tests with special setups" in t/README.

When I added GETTEXT_POISON in bb946bb ("i18n: add GETTEXT_POISON
to simulate unfriendly translator", 2011-02-22) I was concerned with
ensuring that the _() function would get constant folded if NO_GETTEXT
was defined, and likewise that GETTEXT_POISON would be compiled out
unless it was defined.

But as the benchmark in my [1] shows doing a one-off runtime
getenv("GIT_TEST_[...]") is trivial, and since GETTEXT_POISON was
originally added the GIT_TEST_* env variables have become the common
idiom for turning on special test setups.

So change GETTEXT_POISON to work the same way. Now the
GETTEXT_POISON=YesPlease compile-time option is gone, and running the
tests with GIT_TEST_GETTEXT_POISON=[YesPlease|] can be toggled on/off
without recompiling.

This allows for conditionally amending tests to test with/without
poison, similar to what 859fdc0 ("commit-graph: define
GIT_TEST_COMMIT_GRAPH", 2018-08-29) did for GIT_TEST_COMMIT_GRAPH. Do
some of that, now we e.g. always run the t0205-gettext-poison.sh test.

I did enough there to remove the GETTEXT_POISON prerequisite, but its
inverse C_LOCALE_OUTPUT is still around, and surely some tests using
it can be converted to e.g. always set GIT_TEST_GETTEXT_POISON=.

Notes on the implementation:

 * We still compile a dedicated GETTEXT_POISON build in Travis
   CI. Perhaps this should be revisited and integrated into the
   "linux-gcc" build, see ae59a4e ("travis: run tests with
   GIT_TEST_SPLIT_INDEX", 2018-01-07) for prior art in that area. Then
   again maybe not, see [2].

 * We now skip a test in t0000-basic.sh under
   GIT_TEST_GETTEXT_POISON=YesPlease that wasn't skipped before. This
   test relies on C locale output, but due to an edge case in how the
   previous implementation of GETTEXT_POISON worked (reading it from
   GIT-BUILD-OPTIONS) wasn't enabling poison correctly. Now it does,
   and needs to be skipped.

 * The getenv() function is not reentrant, so out of paranoia about
   code of the form:

       printf(_("%s"), getenv("some-env"));

   call use_gettext_poison() in our early setup in git_setup_gettext()
   so we populate the "poison_requested" variable in a codepath that's
   won't suffer from that race condition.

 * We error out in the Makefile if you're still saying
   GETTEXT_POISON=YesPlease to prompt users to change their
   invocation.

 * We should not print out poisoned messages during the test
   initialization itself to keep it more readable, so the test library
   hides the variable if set in $GIT_TEST_GETTEXT_POISON_ORIG during
   setup. See [3].

See also [4] for more on the motivation behind this patch, and the
history of the GETTEXT_POISON facility.

1. https://public-inbox.org/git/[email protected]/
2. https://public-inbox.org/git/[email protected]/
3. https://public-inbox.org/git/[email protected]/
4. https://public-inbox.org/git/[email protected]/

Signed-off-by: Ævar Arnfjörð Bjarmason <[email protected]>
Signed-off-by: Junio C Hamano <[email protected]>

Author: avar

.travis.yml

@@ -26,7 +26,7 @@ addons:
 
 matrix:
   include:
-    - env: jobname=GETTEXT_POISON
+    - env: jobname=GIT_TEST_GETTEXT_POISON
       os: linux
       compiler:
       addons:

Makefile

@@ -362,11 +362,6 @@ all::
 # Define HAVE_DEV_TTY if your system can open /dev/tty to interact with the
 # user.
 #
-# Define GETTEXT_POISON if you are debugging the choice of strings marked
-# for translation.  In a GETTEXT_POISON build, you can turn all strings marked
-# for translation into gibberish by setting the GIT_GETTEXT_POISON variable
-# (to any value) in your environment.
-#
 # Define JSMIN to point to JavaScript minifier that functions as
 # a filter to have gitweb.js minified.
 #
@@ -1452,7 +1447,7 @@ ifdef NO_SYMLINK_HEAD
 	BASIC_CFLAGS += -DNO_SYMLINK_HEAD
 endif
 ifdef GETTEXT_POISON
-	BASIC_CFLAGS += -DGETTEXT_POISON
+$(error The GETTEXT_POISON option has been removed in favor of runtime GIT_TEST_GETTEXT_POISON. See t/README!)
 endif
 ifdef NO_GETTEXT
 	BASIC_CFLAGS += -DNO_GETTEXT
@@ -2603,7 +2598,6 @@ ifdef GIT_TEST_CMP_USE_COPIED_CONTEXT
 	@echo GIT_TEST_CMP_USE_COPIED_CONTEXT=YesPlease >>$@+
 endif
 	@echo NO_GETTEXT=\''$(subst ','\'',$(subst ','\'',$(NO_GETTEXT)))'\' >>$@+
-	@echo GETTEXT_POISON=\''$(subst ','\'',$(subst ','\'',$(GETTEXT_POISON)))'\' >>$@+
 ifdef GIT_PERF_REPEAT_COUNT
 	@echo GIT_PERF_REPEAT_COUNT=\''$(subst ','\'',$(subst ','\'',$(GIT_PERF_REPEAT_COUNT)))'\' >>$@+
 endif

ci/lib-travisci.sh

@@ -123,7 +123,7 @@ osx-clang|osx-gcc)
 	# Travis CI OS X
 	export GIT_SKIP_TESTS="t9810 t9816"
 	;;
-GETTEXT_POISON)
-	export GETTEXT_POISON=YesPlease
+GIT_TEST_GETTEXT_POISON)
+	export GIT_TEST_GETTEXT_POISON=YesPlease
 	;;
 esac

gettext.c

@@ -7,6 +7,7 @@
 #include "gettext.h"
 #include "strbuf.h"
 #include "utf8.h"
+#include "config.h"
 
 #ifndef NO_GETTEXT
 #	include <locale.h>
@@ -46,15 +47,15 @@ const char *get_preferred_languages(void)
 	return NULL;
 }
 
-#ifdef GETTEXT_POISON
 int use_gettext_poison(void)
 {
 	static int poison_requested = -1;
-	if (poison_requested == -1)
-		poison_requested = getenv("GIT_GETTEXT_POISON") ? 1 : 0;
+	if (poison_requested == -1) {
+		const char *v = getenv("GIT_TEST_GETTEXT_POISON");
+		poison_requested = v && strlen(v) ? 1 : 0;
+	}
 	return poison_requested;
 }
-#endif
 
 #ifndef NO_GETTEXT
 static int test_vsnprintf(const char *fmt, ...)
@@ -164,6 +165,8 @@ void git_setup_gettext(void)
 	if (!podir)
 		podir = p = system_path(GIT_LOCALE_PATH);
 
+	use_gettext_poison(); /* getenv() reentrancy paranoia */
+
 	if (!is_directory(podir)) {
 		free(p);
 		return;

gettext.h

@@ -28,25 +28,22 @@
 
 #define FORMAT_PRESERVING(n) __attribute__((format_arg(n)))
 
+extern int use_gettext_poison(void);
+
 #ifndef NO_GETTEXT
 extern void git_setup_gettext(void);
 extern int gettext_width(const char *s);
 #else
 static inline void git_setup_gettext(void)
 {
+	use_gettext_poison(); /* getenv() reentrancy paranoia */
 }
 static inline int gettext_width(const char *s)
 {
 	return strlen(s);
 }
 #endif
 
-#ifdef GETTEXT_POISON
-extern int use_gettext_poison(void);
-#else
-#define use_gettext_poison() 0
-#endif
-
 static inline FORMAT_PRESERVING(1) const char *_(const char *msgid)
 {
 	if (!*msgid)

git-sh-i18n.sh

@@ -17,7 +17,7 @@ export TEXTDOMAINDIR
 
 # First decide what scheme to use...
 GIT_INTERNAL_GETTEXT_SH_SCHEME=fallthrough
-if test -n "$GIT_GETTEXT_POISON"
+if test -n "$GIT_TEST_GETTEXT_POISON"
 then
 	GIT_INTERNAL_GETTEXT_SH_SCHEME=poison
 elif test -n "@@USE_GETTEXT_SCHEME@@"

po/README

@@ -289,16 +289,11 @@ something in the test suite might still depend on the US English
 version of the strings, e.g. to grep some error message or other
 output.
 
-To smoke out issues like these Git can be compiled with gettext poison
-support, at the top-level:
+To smoke out issues like these, Git tested with a translation mode that
+emits gibberish on every call to gettext. To use it run the test suite
+with it, e.g.:
 
-    make GETTEXT_POISON=YesPlease
-
-That'll give you a git which emits gibberish on every call to
-gettext. It's obviously not meant to be installed, but you should run
-the test suite with it:
-
-    cd t && prove -j 9 ./t[0-9]*.sh
+    cd t && GIT_TEST_GETTEXT_POISON=YesPlease prove -j 9 ./t[0-9]*.sh
 
 If tests break with it you should inspect them manually and see if
 what you're translating is sane, i.e. that you're not translating

t/README

@@ -301,6 +301,12 @@ that cannot be easily covered by a few specific test cases. These
 could be enabled by running the test suite with correct GIT_TEST_
 environment set.
 
+GIT_TEST_GETTEXT_POISON=<non-empty?> turns all strings marked for
+translation into gibberish if non-empty (think "test -n"). Used for
+spotting those tests that need to be marked with a C_LOCALE_OUTPUT
+prerequisite when adding more strings for translation. See "Testing
+marked strings" in po/README for details.
+
 GIT_TEST_SPLIT_INDEX=<boolean> forces split-index mode on the whole
 test suite. Accept any boolean values that are accepted by git-config.
 

t/lib-gettext.sh

@@ -12,7 +12,7 @@ export GIT_TEXTDOMAINDIR GIT_PO_PATH
 
 . "$GIT_BUILD_DIR"/git-sh-i18n
 
-if test_have_prereq GETTEXT && ! test_have_prereq GETTEXT_POISON
+if test_have_prereq GETTEXT && test_have_prereq C_LOCALE_OUTPUT
 then
 	# is_IS.UTF-8 on Solaris and FreeBSD, is_IS.utf8 on Debian
 	is_IS_locale=$(locale -a 2>/dev/null |

t/t0000-basic.sh

@@ -274,7 +274,7 @@ test_expect_success 'pretend we have a mix of all possible results' "
 	EOF
 "
 
-test_expect_success 'test --verbose' '
+test_expect_success C_LOCALE_OUTPUT 'test --verbose' '
 	test_must_fail run_sub_test_lib_test \
 		test-verbose "test verbose" --verbose <<-\EOF &&
 	test_expect_success "passing test" true